1. Rhythmic activity was induced with either serotonin (5-HT; 10-100 microM) or dopamine (0.1-1.0 mM), in the in vitro spinal cord preparation of neonatal rats, with one intact hindlimb attached. Patterns of activity were investigated with multiple EMG recordings and the spatiotemporal characteristics of 5-HT- and dopamine-induced activity compared. 2. Dopamine-induced rhythmic activity was slow (cycle duration: 2.2-70.1 s) and irregular, whereas rhythmic activity induced by 5-HT was fast (cycle duration: 1.3-5.1 s) and regular. 3. During 5-HT- and dopamine-induced rhythmic activity, the timing of muscular activity was similar for hip flexors and hip adductors, for semimembranosus (hip extensor), and for muscles controlling the ankle and the foot. 4. In contrast, notable differences in the phase in the pattern induced by 5-HT compared with that induced by dopamine were found in the biceps femoris, semitendinosus, and quadriceps muscles. Biceps femoris and semitendinosus (functional hip extensors and knee flexors) were always extensor-like during 5-HT-induced activity, whereas in dopamine, these muscles displayed flexor-like bursts and double bursts as well as extensor-like bursts. Lack of EMG activity in biceps femoris and semitendinosus was encountered also in dopamine. In rectus femoris, vastus lateralis, and vastus medialis (main function: knee extension), the activity was dominated by flexor-like bursts in 5-HT, whereas in dopamine the activity was shifted to a predominantly extensor-like pattern. 5. The relationship between flexor and extensor burst duration and cycle duration was more variable than described for locomotor activity in adult animals. 6. The relative timing of muscle activity was stable from P0 to P4. The most important difference between rats aged 0-1 days and rats aged 2-4 days was a delayed flexor-extensor transition in older animals. 7. The complex timing of hindlimb muscle activity was relatively unchanged after transecting all dorsal roots. 8. Finally, the relationship between flexor and extensor activity and ventral root discharges was determined. It was found that the L2 ventral root burst was in phase with simple flexors while the L5 burst coincide with the extensor phase. 9. We conclude, that 5-HT and dopamine can activate spinal central pattern generators (CPGs) that already at birth are able to produce distinct patterns of motor activity. Modulatory inputs thus seems to be able to reconfigure the CPGs to produce specific motor outputs.